This research aims to examine which ratio between PER, PEG, and PERG that can better predict the stock returns of the firms, in addition to analyze the effect of each that ratio onstock return. The research sample consist of 310 non-financial firms in IDX during the period of 2010-2016. In this study, multiple linear regression methods have been conducted to explain the effect of PER, PEG, and PERG with control variable SIZE, M/B, FLEV, and DPRon stock return. The results indicate that PERG can explain stock returns better than PER and PEG based on the higher value of R-square. Using 5% level of significance, M/B and FLEV had a significant effect, while PER, PEG, and PERG had no significant effect on stock return. Thestudy also showed a negative effect PER, PEG, PERG, SIZE, FLEV, and DPR as well as thepositive effect of M/B on stock returns.